Predicting the effects of landuse on water quality – Stage I
Pollution Risk Modelling (Landcare Research, Objective 5)
EnSus is a framework for analyzing and mapping the relative risks different land uses pose to soil quality and water quality. EnSus has been used to map relative risk classes of nitrate leakage from soils to surface and ground waters. It uses best available knowledge of specified land use pressures and vulnerability of the land to those pressures.
EnSus complements the national SPARROW modelling work (Section 5.2) for N and P. However the EnSus approach is at finer spatial scales than SPARROW, and does not estimate spatially integrated responses over catchments, or take account of in-stream attenuation processes. The EnSus model can be summarised as a set of rules that combine maps of soil attributes, rainfall, and land use/management into maps of leaching risk. These rules are documented in this section, and can easily be implemented as part of the catchment modelling framework.
The process involved three steps:
• mapping vulnerability of soils to N leaching from the soil,
• mapping land use as an estimate of N input pressure, and
• combining vulnerability and pressure to estimate risk.
Risk maps are provided for the South Island, North Island (200m raster), and Waikato lowlands. The South Island and North Island maps are intended for large catchment, regional, and national applications. More detailed applications will require analysis based on available higher resolution soil maps. The Waikato lowlands map is provided to show the results that may be obtained from a higher resolution soil map.
Vulnerability to Leaching
Three outputs are produced
1. Potential N leaching index for nitrate mobilised from the soil that is likely to contribute to either ground water or surface water N runoff.
2. Likely attenuation of nitrate on route to water bodies, by passage through wet, reduced soils.
3. Intersection of 1 and 2 as an indicator of relative risk of nitrate leaching to water bodies.
Potential N leaching index
Potential leaching was estimated using the Land Environments of New Zealand national layer of rainfall to evaporation ratio (RF/ET) based on Meteorological Service monthly data modelled as a mean annual national surface. This ratio was modified (1) by a ‘PAW Factor’ used to increase the index where profile available water (PAW) is lower than 200mm (to account for extra leaching in low PAW soils), and (2) by a ‘slow permeability factor’ used to decrease the index where permeability is very slow (to account for loss of potential leaching water as runoff).
The potential leaching index was calculated as (RF/ET) * (PAW Factor) * (Slow permeability factor). This estimates the relative potential for N mobilisation from the soil (without specifying if this is mobilised to surface or ground waters.).
The PAW Factor was determined by the relationship between the water surplus modelled and reported by Met Service, and the benchmark PAW values (40, 80, 120 and 160mm water storage). The PAW multipliers in Table 9-1 are provided for soils under mean long term average rainfall of 1000mm or more, and less than 1000mm. It is assumed that there is an insignificant effect of PAW on relative leaching, when PAW exceeds 200mm.
Table 9-1: Factors to calculate the potential leaching by increasing effective rainfall where PAW is less than 200mm
| PAW | PAW multiplier Rainfall > or = 1000mm |
PAW multiplier Rainfall <1000mm |
|---|---|---|
| <40 | 1.4 | 2.4 |
| 40 - 69 | 1.3 | 2.1 |
| 70 - 99 | 1.2 | 1.8 |
| 100 – 199 | 1.1 | 1.4 |
| >200 | 1 | 1 |
Soils with very slow permeability (saturated hydraulic conductivity <2.5 mm/day) were identified in the NZLRI soil legend. For these soils, the potential leaching index was reduced by a factor of 30%.
Attenuation of N via pathway to water bodies
Attenuation is defined here as denitrification and loss of nitrogen to the atmosphere as either nitrous oxide or nitrogen gas. It is assumed that nitrogen is primarily in the form of nitrate. The attenuation layer is an independent layer that may be used to reduce the potential leaching index and provide an estimate of the attenuation of nitrate by passage through soils periodically saturated with water.
The effect of tile or mole drains where potential attenuation is bypassed is not considered. Attenuation is predicted by two means:
1. Presence of Gley Soils, Organic Soils and imperfectly drained soils that have very slow saturated hydraulic conductivity (less than 2.5 mm/day). These are based on the soil theme of the NZLRI.
2. Presence of soil associations where Gley or Organic Soils are likely to occur as riparian strips but are too small to be shown on soil maps. These areas were identified by delineating land systems, based on NZLRI land units, in which well expressed drainage catenas were likely to occur.
The attenuation effect is expressed as a multiplier in Table 9-2.
Table 9-2: Combination of attenuation by soil class and riparian class (a small multiplier indicates that leaching is greatly reduced by that drainage class).
| Attenuating Drainage class | Multiplier |
|---|---|
| Very poorly drained (Organic Soils) | 0.01 |
| Poorly drained (Gley orders, groups and subgroups) | 0.5 |
| Peaty-gley subgroups | 0.2 |
| For remainder, Land with riparian Gley soils | 0.5 |
| Imperfectly drained & very slowly permeable soils in land with likely riparian Gley soils | 0.7 |
| Imperfectly drained & very slowly permeable soils in land without riparian Gley soils | 0.8 |
| For remainder, land with likely riparian Gley soils | 0.8 |
| For remainder | 1 |
Vulnerability classes
The potential N leaching index which ranges from 0 – 44, was divided into 5 classes with the limits: 0, 2, 3, 4, 7, 44. These limits best express our understanding of potential leaching contrasts across the soil-landform-rainfall pattern. The scale is not linear and strongly influenced by effective rainfall. Class 5 (7 – 44) is mainly confined to mountainous regions with high rainfall.
Pressure
The pressure of N inputs to soils was estimated from land use classes based on Agribase and with the addition of LCDB1 to fill in the gaps. Agribase and LCDB1 categories were combined into nine land use classes (see column 1 of Table 9-3). N pressure index was assigned to the land use classes. This was estimated based on knowledge of N (kg/ha/yr) leached from land uses at a relatively low number of sites.
Dairying was divided into 2 classes; high intensity (>200 cows/farm) and moderate intensity (<= 200 cows/farm), based on stock numbers from Agribase.
Risk of nitrate leaching
In this analysis relative risk is derived from the combination of pressure on vulnerability. We do not consider sensitivity or asset value in this analysis. Only one hazard, N leaching from the soil, is considered. Vulnerability and pressure are combined in Table 9-3. It produces 3 classes of risk, but can be modified to provide more classes.
Table 9-3: Combination of N leaching vulnerability and N pressure to derive relative risk, where risk = vulnerability index * pressure index, with risk classes: very low <3, low = 3-7, mod = 8-16, high = 17-29, very high 30-50.
| Land use class and N pressure index | N leaching vulnerability | |||||
|---|---|---|---|---|---|---|
| Low (1) |
Mod low (2) |
Mod (3) |
Mod High (4) |
H (5) | ||
| ARA arable | 10 | 10 | 20 | 30 | 40 | 50 |
| DAI2 dairy >200 | 10 | 10 | 20 | 30 | 40 | 50 |
| DAI1 dairy 0-200 | 8 | 8 | 16 | 24 | 32 | 40 |
| SBO sheep beef + | 3 | 3 | 6 | 8 | 10 | 12 |
| NAR non arable | 2 | 2 | 4 | 6 | 8 | 10 |
| FOR exotic forest | 1 | 1 | 2 | 3 | 4 | 5 |
| NAT native | 0.5 | 0.5 | 1 | 1.5 | 2 | 2.5 |
| TUSS tussock | 0.5 | 0.5 | 1 | 1.5 | 2 | 2.5 |
| ARTIF urban etc. | 0.5 | 0.5 | 1 | 1.5 | 2 | 2.5 |
Uncertainty in this analysis is introduced by:
1. Accuracy of the index of mean annual rainfall to evapotranspiration layer and its applicability as an index of potential leaching.
2. The appropriateness of multipliers for PAW, very slow permeability, and attenuation in wet soils.
3. Accuracy of soil map representations of PAW, very slow permeability soils, and wet reduced soil layers including identification of land units with poorly drained riparian strips.
4. Choice of vulnerability classes.
5. Combination of Agribase land use categories, and estimation of N pressure index.
6. Method for combination of pressure and vulnerability, and choice of risk class limits.
It is not possible to express the sensitivity of the result to these uncertainties without further analysis. Use of more detailed scale soil maps where available, will substantially decrease uncertainties in category 3.
Results for Nitrate Leaching
Maps of relative risk of nitrate leaching are shown for the North Island and South Island in Figure 9-1 and Figure 9-2, respectively. Risk is expressed in the 5 classes of Table 9-3. The maximum resolution of the data is 200m. A smaller map indicates uncertainty in the underlying soil data.
The data used to generate the risk maps is available for the following layers:
1. Potential N leaching index (PNLI)
2. PNLI modified by attenuation in Gley and Organic Soils (Ren1)
3. PNLI modified by attenuation in riparian Gley and Organic Soils (Ren2)
4. PNLI modified by combined Ren1 and Ren2
5. Nitrate leaching risk based on land use pressure and PNLI modified by combined Ren1 and Ren2
A nitrate leaching risk map of Waikato lowlands (Figure 9-3) shows the five risk classes (Table 9-3) at 25m resolution. It is based on more detailed soil maps than the soil data underpinning the national maps. Comparison is provided with the same area clipped from the 200m-resolution national map. Compared to the 200m map, the more detailed 25m map shows finer scale patterns of risk (for example lower risk along stream lines in the southern end of the area) and changes in risk due to improved soil data, particularly in areas of peat soils. Greater contrast between risk information based on the older national map and that based on recent more detailed maps would be expected in other parts of New Zealand where resurvey has radically upgraded the older information.
Figure 9-1: Relative risk of nitrate leaching for the North Island
Figure 9-2: Relative risk of nitrate leaching for the South Island
Figure 9-3: Relative risk of nitrate leaching for the Waikato, showing the effect of improved soils data.
Contact for Enquiries
MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND
Fax: +64 4 894 0721
Contact this person
